一种线性自抗扰控制器参数自整定方法

针对线性自抗扰控制器参数难于整定的问题,提出了一种基于动态响应过程时序数据挖掘的参数自整定算法. 算法以线性自抗扰控制器中线性误差反馈律的两个增益信号回路的动态响应为参数调整对象,通过改进变收缩系数的随机搜索算法进行参数整定,记录动态响应过程数据,基于关联关系挖掘得到控制参数调整策略应用于线性自抗扰控制器的参数自整定. 为验证本文提出的参数自整定方法的实际效果,以液压自动位置控制系统为控制对象,分别采用阶跃响应仿真和Monte Carlo实验进行对比研究. 结果表明,基于数据挖掘参数自整定的线性自抗扰控制器动态响应较好,鲁棒性较强,改进了变收缩系数随机搜索算法调整时间较长以及传统线性自抗扰控制器超调较大的缺点,是一种具有实用性的线性自抗扰控制器参数自整定方法.

Self-tuning method for a linear active disturbance rejection controller

This paper introduces a parameter self-tuning algorithm based on dynamic response time series data mining to solve the parameter self-tuning difficulty of a linear active disturbance rejection controller ( LADRC) . In the algorithm, two gain signal loops' dynamic response of linear state error feedback ( LSEF ) is used as the parameter adjustment object. Through the parameters auto-tuned by NLJ algorithm and the process data recorded, the control parameter adjustment policy based on association mining is ap-plied to LADRC parameter auto-tuning. To verify the actual effect of the parameter self-tuning method in this paper, a hydraulic automatic position control ( HAPC) system is used as the control object. Step response simulation and Monte Carlo experiment show that the dynamic response of the system which is combined by HAPC and the controller is better, the robustness is stronger, the adjust-ment time is shorter than NLJ algorithm, and the overshoot is also less than the traditional LADRC controller. It is considered as a practical LADRC controller parameter self-tuning method.